Just made my first attempt at cooking using liquid nitrogen. Decided to start easy, and make a 0.5L batch of ice cream, along with some powdered peanut-butter and banana for toppings. Overall I'd rate the entire experiment a success, but noticed something a bit strange when creating the powdered elements.

My technique was roughly as follows:

  1. Place a bit of banana and a dollop of peanut butter in a small glass bowl with LN2, and leave until the 'boiling' mostly subsides (replenishing as needed; this probably took about 400mL of LN2).

  2. Prechill a mortar and pestle with LN2, then transfer the frozen banana and peanut-butter to it (I did each one separately, though probably it would have been fine to do both together), along with enough LN2 to keep them nice and cold (probably about 300mL of LN2 to 'prime' the mortar and pestle to the point where it was holding a nice shallow pool of LN2).

  3. Crush the frozen item(s) down to the desired consistency, replenishing LN2 as needed (this is where I needed the most LN2, I lost track of how much I was using but may have gone as high as 1L per item). I was going for 'fine dust' but had to stop a bit before reaching that point.

The strangeness was observed during #3. Basically what I found was the more finely I crushed the frozen food, the faster the LN2 would evaporate off, and the more I'd need to put in to keep everything nice and cold. I wasn't grinding particularly aggressively, and towards the end it got to the point where the LN2 would disappear only a second or two after pouring it in, making grinding effectively impossible (unless I wanted to risk ending up with thawed mush, I suppose). This ultimately ended up consuming more LN2 than it had taken to chill the ice cream, even though in terms of volume there was far more ice cream than banana or peanut-butter.

Is there a reason/explanation for this? Maybe something to do with the effective surface area increasing as the frozen food is ground into smaller and smaller pieces? Or was my technique wrong? Is there a better way to turn arbitrary things into frozen powders using LN2?


3 Answers 3


Your 4 observations:

  1. Quick evaporation first showed up when finely divided particles entered into the situation
  2. The finer the frozen food was crushed the faster the evaporation
  3. Not grinding particularly aggressively (i.e. probably friction-heat from grinding not generated faster as the experiment progressed)
  4. Toward the end of the experiment evaporation completed before any grinding even took place

sound like your speculation: "Maybe something to do with the effective surface area increasing as the frozen food is ground into smaller and smaller pieces?" is on the right track.

The particles of frozen banana & peanut butter provide lots of places for the boiling of the LN2 to occur, just like if you dump a bunch of salt into almost-boiling water, it suddenly boils up like crazy where the grains hit the water.


Grinding items into smaller pieces creates far more surface area per total volume. With heat transfer, a key component in how fast the transfer will occur is surface area, and by providing more you allowed the transfer to occur faster.

Think of a couple examples, if you use ice chips in a drink as compared to one large cube, the drink will become cold faster, but the ice will also melt faster. If you have a hot liquid you want to cool, keeping it in a bowl will be much slower than dumping it onto a cookie sheet and letting it spread for more surface area.


You're putting a fair bit of heat in with all that effort. This then has to be added to the heat conducted in from the surroundings. I've made LN ice cream a few times, using no more than 2 litres of LN to make over a litre of ice cream, so it feels like you got through too much (it sounds like you got the same impression)

I'm not sure what the point of grinding frozen peanut butter is - perhaps you could use finely ground peanuts as a topping and/or stir smooth peanut butter into your liquid ingredients. Similarly putting the banana into the mix would take less cooling. If you want to serve LN-frozen banana powder you need to keep it frozen until served, which might not be easy (I reckon it would stick together in the freezer). I suggest you'd be better off serving finely-chopped banana, which you could do starting from freezer-cold, then only keep it cold with the LN.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.